For several decades, Australian higher education policymakers have been interested in the idea that there is a desirable ‘balance’ between public and private contributions to the cost of higher education, and that a distribution of public and private benefits should inform this.

In an earlier post, I argued that a Deloitte Access Economics report released this week had come closer than any previous work to calculating a distribution of public and private benefits of higher education. What I am not convinced of is that such a calculation is useful for policymakers.

Sometimes an analysis of personal benefits and public benefits, as distinct from some ‘balance’ between them, will be helpful. In the Deloitte report (p.10) they argue that:

The economic policy rationale for governments to support higher education is the existence of a ‘market failure’ – specifically, the existence of the public benefits described above and the fact that, in the absence of government funding, the decisions by providers and students will not drive the system toward its socially optimal operation.

Economic theory suggests that students will choose to acquire knowledge where their expected private benefit is at least equal to their cost of education. If at least some public benefit exists, then this decision-making process will result in a suboptimal level of knowledge transfer activities.

In order to increase levels of knowledge and maximise the total net social benefit of higher education, governments need to be able to identify the public benefits being created, such that appropriate subsidies can be derived and applied. Identifying the relative split between public and private benefits may then inform the relative subsidy payments based on these dimensions.

Apart from the sentences in bold, I agree. I have made similar arguments myself.

The problem with the first bolded sentence is that the presence of public benefits does not of itself lead to sub-optimal levels of education. This will only happen if the total net private benefits are too low to justify enrolment. In those cases, tuition subsidies reduce costs and make it easier to get to positive net private benefits. This may encourage prospective students to enrol when otherwise they would not.

The main argument of my 2012 Graduate Winners report is that even though market failures are possible, with income contingent loans there are only limited empirical circumstances in which they actually exist.

In most cases the private benefits of higher education are already so large – Deloitte, like previous research, identifies hundreds of thousands of dollars or more extra in lifetime income (p.34) – that the tuition subsidies are unlikely to sway the decisions of someone acting in their rational economic self-interest. Subsidies at the levels historically seen in Australia usually add relatively small amounts to net private financial benefits that are already large enough to attract students to higher education. And this is before we take into account other factors influencing people to attend higher education, such as interest in their field of study, access to particular careers, the lifestyle experience of campus, status, and keeping parents happy.

Of course, people don’t always act in their rational economic self-interest. But demand for higher education has rarely been the policy problem we face, which over the last few decades has generally been an under-supply of student places (which, paradoxically, public funding can exacerbate).

The English experience of this is interesting. As noted in another post, they have abolished most direct tuition subsidies and charge above-cost fees for some courses. Despite this, in the school leaver population demand is at record levels as a share of population, possibly because their other options are even worse than paying high university fees. The story is different for mature-age students, highlighting the importance of empirical analysis of demand in policy.

The more compelling market failure argument is in the capital markets. School leaver age students are unlikely to have the money to pay for their own university education, and their parents may be unable or unwilling to pay for them. Banks are reluctant to lend for education and charge high interest rates when they do, due to the lack of assets to recover if a student defaults. Tuition subsidies are one way of overcoming this market failure. But government-backed loans, and especially income contingent loans, are another way of avoiding capital market failure in higher education that has proven to be successful.

But if the government does decide that demand is too low, despite loans being available, an analysis of private and public benefits could be useful in deciding where to invest their money in attracting more students to achieve the greatest public return on investment. But a public-private benefits model based primarily on the economy, as the Deloitte model is, won’t necessarily be the right analytical approach. At the discipline level, public and private financial benefits tend to be correlated because the broader public productivity effects are usually also reflected in wages. But this means that the private benefits should already be fairly high in most cases of high potential public benefits, and enhancing already high private benefits with additional tuition subsidies probably won’t make a big difference to prospective student decision making.

The situation most likely to trigger market failure is low private benefits but high public benefits, but those public benefits are often going to require non-economic analysis. Possible examples are occupations with client groups who are not wealthy enough to deliver high private financial returns, and/or the government as a dominant employer that keeps wages down. Nursing and teaching might be examples of jobs in which people believe the public benefits of their work is not fully reflected in salaries; although the Deloitte report shows that nursing has strong private benefits.

Even in these cases, adding up benefits broadly defined and dividing them between public and private is not going to tell us much about how to allocate the cost of providing the course. The prospective student is looking at their own personal costs versus benefits, and it may be that the subsidy required to persuade them to enrol is a zero, small, medium, or large percentage of the total costs of providing the course; or a subsidy of over 100 per cent of course costs might be needed to make the personal economics work, through a scholarship perhaps.

In market failure theory, the government should never pay tuition subsidies exceeding the expected public benefits, but that amount needed to change behaviour could be a wide range of percentages of course costs. The necessary subsidy may also differ between individuals, as we are seeing in the English demographic differences.

The public-private balance idea isn’t one that can be derived from market failure economics. Instead, it is one of the fairness arguments used in higher education politics. I think it came about partly because the arguments for HECS created a logical trap: if students should pay their benefits, why shouldn’t the public pay for its benefits? But there are other ways of thinking about how a student pricing system should work, and we have tended to use them simultaneously without recognising their tensions.

Since 1997, HECS/student contributions have had an assumed relationship to future earnings, with student contribution levels ranked (very roughly) by potential future private financial benefits. This links to both market ideas (pay more, get more) and to the progressive notions embedded in the Australian tax and welfare system (the rich should pay more and get less).

The total funding rate for a course is loosely based on course delivery costs: low funding rates for chalk and talk (or these days Powerpoint and pointer) courses like arts, law or commerce but high funding rates for courses with laboratories or high equipment costs such as engineering, science or medicine, with others in between.

In high private benefit/low course delivery cost disciplines like law students pay most of the total costs (84 per cent), while in low private benefit/moderate course cost fields like performing arts students pay only 33 per cent of the costs. These different percentages of costs are often seen as anomalies, but they are the logical consequence of private benefits and course costs being independent of each other. There is not supposed to be any ‘balance’ between public and private payment when the key variables are future earnings and course delivery costs.

A funding formula based on private benefits as a percentage of total benefits applied to course costs (the long-pursued ‘balance’), as opposed to one based largely on future-earnings ranked prices (broadly the current system), leads to results that are politically difficult.

On the Deloitte figures, under a public-private balance system nurses should be paying 50 per cent of their costs, not their current 33 per cent. Law and business students should paying 44 per cent of their costs, well below their current 84 per cent.

Nurses paying more while lawyers and accountants pay less does not strike me as a winning political proposition. This is not just because public opinion has a status quo bias. It integrates poorly with the fairness ideas present in other aspects of Australia’s tax and welfare system.

A rigid division of public and private costs based on a calculation of public and private benefits would also give policymakers less room to respond to real market failures. These will show in recent evidence such as skills shortages and weak demand from prospective students, not benefit analysis based on historical data.

The public and private benefits of higher education remain interesting topics in themselves. In certain contexts they can, separately, guide policy. But the idea that there is a ‘balance’ between them that ought to be reflected in funding policy is not sound. There are better ways of thinking about how to pay for higher education.